Things we must solve before V29

I don't quite know where Koshling is on the disappearing unit issue. I do not know if that has anything to do with the bug I had in my clumsy fix or not. Once he updates, we need to test a bit because that bug was extremely elusive to duplicate since a save and reload seems to fix it for some strange reason (temporarily).

So... Koshling, have you looked at this one yet? I've made no progress on even getting a theory what could be going wrong there if it has nothing to do with my bad fix... The long days at work keep me from making much headway on that.

Additionally, are there any other bugs anyone else knows about that are yet to be addressed for THIS cycle? (I don't know if you'll want to address the two my wife are posting tonight before release, though that scout one is pretty rough bad AI decision making!)
 
Just the opinion of a junkie like me, but this disappearing bug sounds bad, so I think it would be better to wait a bit and be sure it's fixed.

Also patiently waiting for a release, keep up the good job :)
 
Update on Combat Modifier problem

Ok, did some digging and ran a testgame though I'm out of time to do a better eval and I have a feeling you're going to find this a lot faster than I can. This one really baffles me actually.

This is the missing Combat Class modifier problem but on taking a closer look, its actually odder than that. Things just flat out aren't calculating correctly here. If you find the rogue he shows in the pic in this post and select it you can see that it goes further than just the combat evaluation... the unit details itself don't include the modifier vs melee and most of those are somehow out of calculatory whack (and this is after a recalc!!)

And yet, on my wife's game, she can show me examples where this is NOT taking place!

Ok... I don't think my fix would've resolved this at all because this is a problem in the way the totals are tallying completely but HOW that would become a problem in the first place is outside of any theory I currently have to offer.

Anyhow, here's the post with a save that clearly expresses the issue.

Problem confirmed on the most recent SVN updates so its still a major issue at large and is definitely an indicator of a fairly serious issue at that.
 
I don't quite know where Koshling is on the disappearing unit issue. I do not know if that has anything to do with the bug I had in my clumsy fix or not. Once he updates, we need to test a bit because that bug was extremely elusive to duplicate since a save and reload seems to fix it for some strange reason (temporarily).

So... Koshling, have you looked at this one yet? I've made no progress on even getting a theory what could be going wrong there if it has nothing to do with my bad fix... The long days at work keep me from making much headway on that.
I'm not looking at this one at all. I can't look t everything, and I'm getting a bit stressed out at the current volume. I have no plans to look at this one.

Additionally, are there any other bugs anyone else knows about that are yet to be addressed for THIS cycle? (I don't know if you'll want to address the two my wife are posting tonight before release, though that scout one is pretty rough bad AI decision making!)
Is that the one where the Mauri are just building scouts? If so I plan to look at it today

Update on Combat Modifier problem

Ok, did some digging and ran a testgame though I'm out of time to do a better eval and I have a feeling you're going to find this a lot faster than I can. This one really baffles me actually.

This is the missing Combat Class modifier problem but on taking a closer look, its actually odder than that. Things just flat out aren't calculating correctly here. If you find the rogue he shows in the pic in this post and select it you can see that it goes further than just the combat evaluation... the unit details itself don't include the modifier vs melee and most of those are somehow out of calculatory whack (and this is after a recalc!!)

And yet, on my wife's game, she can show me examples where this is NOT taking place!

Ok... I don't think my fix would've resolved this at all because this is a problem in the way the totals are tallying completely but HOW that would become a problem in the first place is outside of any theory I currently have to offer.

Anyhow, here's the post with a save that clearly expresses the issue.

Problem confirmed on the most recent SVN updates so its still a major issue at large and is definitely an indicator of a fairly serious issue at that.

Modifier recalc does NOT fix unit modifiers currently, so if old promotion bugs have screwed up a unit's modifiers (entirely possible) a recalc will no fix them. If the bug is explainable through the unit having the wrong modifiers then we should either:
1) Just leave it be unless someone can reproduce it on a NEW unit (thereby ruling out historical promotion issues as a possible source); or
2) We need to add unit modifier recalc to the recalc process and verify that fixes it in this save (but that's more work than I wanted to do on the recalc system really)

Since we KNOW the unit in question was impacted by promotion loss and therefore hit by promotion bug issues I'd say it's most likely just a hold over from that in the saved game state, and we shouldn't regard this as a V29 problem without further evidence.

If someone wants to add unit modifier recalculation I should be a fairly mechanical (but possible not small) job - just do like all the other recalcs on game/player/city - clear all modifiers down to 0 then replay the processing of the promotions.
 
I'm not looking at this one at all. I can't look t everything, and I'm getting a bit stressed out at the current volume. I have no plans to look at this one.
Actually... I don't blame you. I may be able to get my wife's playtesting to help me pinpoint this one a bit better. We've been discussing how to nail it down. If I can at least start a process of nailing it down, I'll let you know if it goes into something I can't figure out. My only concern is that it could have something to do with viewport code which I don't think I'd be all that effective at debugging but at least I can try to help get a bead on the problem. I'll look for it here over the next few days and keep you updated on what we find.


Is that the one where the Mauri are just building scouts? If so I plan to look at it today
Ok. Cool!



Modifier recalc does NOT fix unit modifiers currently, so if old promotion bugs have screwed up a unit's modifiers (entirely possible) a recalc will no fix them. If the bug is explainable through the unit having the wrong modifiers then we should either:
1) Just leave it be unless someone can reproduce it on a NEW unit (thereby ruling out historical promotion issues as a possible source); or
2) We need to add unit modifier recalc to the recalc process and verify that fixes it in this save (but that's more work than I wanted to do on the recalc system really)

Since we KNOW the unit in question was impacted by promotion loss and therefore hit by promotion bug issues I'd say it's most likely just a hold over from that in the saved game state, and we shouldn't regard this as a V29 problem without further evidence.

If someone wants to add unit modifier recalculation I should be a fairly mechanical (but possible not small) job - just do like all the other recalcs on game/player/city - clear all modifiers down to 0 then replay the processing of the promotions.
My thoughts on the matter are pretty much the same here. I doubt the promotion loss issue would have caused this inaccuracy from what I know about that problem. But an issue in the wrapper MAY have... but I don't believe we had that issue then. Furthermore, the source of the problem, as reported, was far stranger, something to do with once they'd made their first attack that utilized such a combat modifier, it would work, then never would again thereafter. After looking at this save I figured it'd be nearly impossible to determine how this is taking place without getting a save that can show how the loss of proper calculation is processing in the first place.

As for the recalculation, I was considering doing just that, but wow would that add some processing time on recalcs or what!?!

It seems not everyone is experiencing this issue so we'll just have to keep a watch out for another report on the problem and try to get those who encounter it to get a save that will show how it happens in the first place. Till then we're pretty much stuck without implementing a unit recalc process which would be, yes, a lengthy effort to implement. Far lengthier than would be worth doing before the next release I'd say.
 
Actually... I don't blame you. I may be able to get my wife's playtesting to help me pinpoint this one a bit better. We've been discussing how to nail it down. If I can at least start a process of nailing it down, I'll let you know if it goes into something I can't figure out. My only concern is that it could have something to do with viewport code which I don't think I'd be all that effective at debugging but at least I can try to help get a bead on the problem. I'll look for it here over the next few days and keep you updated on what we find.



Ok. Cool!




My thoughts on the matter are pretty much the same here. I doubt the promotion loss issue would have caused this inaccuracy from what I know about that problem. But an issue in the wrapper MAY have... but I don't believe we had that issue then. Furthermore, the source of the problem, as reported, was far stranger, something to do with once they'd made their first attack that utilized such a combat modifier, it would work, then never would again thereafter. After looking at this save I figured it'd be nearly impossible to determine how this is taking place without getting a save that can show how the loss of proper calculation is processing in the first place.

As for the recalculation, I was considering doing just that, but wow would that add some processing time on recalcs or what!?!

It seems not everyone is experiencing this issue so we'll just have to keep a watch out for another report on the problem and try to get those who encounter it to get a save that will show how it happens in the first place. Till then we're pretty much stuck without implementing a unit recalc process which would be, yes, a lengthy effort to implement. Far lengthier than would be worth doing before the next release I'd say.

It would add processing time on recalcs, yes, but you don; do recalcs that often, and I think we're talking about 20% extra or something, not many times more.
 
Yeah, might not be as bad as I fear. But in the later game, in a multiplayer connection, recalcs are already long enough to time out the connection in some cases. So it might be beneficial to isolate such a recalc into its own hotkey or into a second phase (but then we might need to eventually put more than just this into a second phase too.)

For now it wouldn't be necessary perhaps. But later on it might be.

I can program that out. I'd be happy to try my hand at it. It's kinda my region of influence anyhow. But again, not right away (aka before release) unless it continues to be the only answer to some issues that come up. Don't want to introduce another potential cause for a bug at this point.
 
Maybe this has already been addressed. I haven't tried them in a while... but if not... what about the bug where loading a savegame with build lists causes the game to crash?
 
Maybe this has already been addressed. I haven't tried them in a while... but if not... what about the bug where loading a savegame with build lists causes the game to crash?

I was not aware of it. Is it a property of the save, or a property of your user settings at the time of the load?

Edit - infinite scout building bug fixed now (see SVN thread)
 
I was not aware of it. Is it a property of the save, or a property of your user settings at the time of the load?

Edit - infinite scout building bug fixed now (see SVN thread)

I think that AIAndy addressed this a while back. At least, I haven't heard it reported in a month or so.

As for the Invisible units and Release, as far as I can tell the problem seems localized to T-brd's wife's computer as I can not get it to reproduce on my machine with the saves he has uploaded or with my own saves. And given that no one else is complaining about it I think we are good to release unless something crazy happens this evening.
 
I think that AIAndy addressed this a while back. At least, I haven't heard it reported in a month or so.

As for the Invisible units and Release, as far as I can tell the problem seems localized to T-brd's wife's computer as I can not get it to reproduce on my machine with the saves he has uploaded or with my own saves. And given that no one else is complaining about it I think we are good to release unless something crazy happens this evening.

It may be related to the invisible forests many people have been reporting. That may be just that somewhere we are not refreshing the screen properly somewhere. The forest problem goes away on save/load and even just going into the World Builder.
 
I'm a little concerned about the exceptions that gave rise to the python errors Joseph was seeing, but without a save tha reproduces it there's not much we can do.

Other than that I think were basically ok - other issues a more (legitimate) disagreeents on tuning than bugs, and at some point we have to draw the line.
 
I'm a little concerned about the exceptions that gave rise to the python errors Joseph was seeing, but without a save tha reproduces it there's not much we can do.

Other than that I think were basically ok - other issues a more (legitimate) disagreeents on tuning than bugs, and at some point we have to draw the line.

I've just committed the mod name change, so at this point I'm uploading unless there is a disaster.
 
I think that AIAndy addressed this a while back. At least, I haven't heard it reported in a month or so.

As for the Invisible units and Release, as far as I can tell the problem seems localized to T-brd's wife's computer as I can not get it to reproduce on my machine with the saves he has uploaded or with my own saves. And given that no one else is complaining about it I think we are good to release unless something crazy happens this evening.

I don't blame you for wanting to proceed. But I warn that this bug is going to end up enforcing the need for a patch. SO experienced this bug as well. It also doesn't just limit the ability of a unit to be seen... it also keeps the unit that would turn invisible when approached, from being attackable as well. And apparently it has not been resolved by the fix to my erroneous attempt at a fix to the other issue.

I believe she's got a save now that makes the issue replicable. Haven't had a chance to look at it as of this posting, but I will take a look tonight.

EDIT: Apparently its not as replicable as we'd hoped. This one's going to be a haunt for a while then apparently. Without having any way to lock it down, I'd say we're clear as we can be for release despite it. (though there was one text reference in DH's stuff I have a feeling he'll have resolved this evening.)
 
Anyone know if invisible trees are still happening or if they were fixed?
 
Back
Top Bottom